Cerebral Cavernous Malformations (CCM) are vascular lesions resulting in seizure and stroke. CCMs result from bi-allelic loss of ccm1, -2, or -3 in endothelial cells; however, the physiological and biochemical mechanisms of lesion development remain incompletely characterized. Loss of CCM1, -2 or -3 protein results in a significant increase in total and active RhoA, along with increased Rho kinase (ROCK) and ROCK effector activity. This biochemical dysregulation is also associated with a number of functional defects, including loss of tube formation, invasion, increased endothelial cell monolayer permeability, and increased actin stress fibers. Significantly, these defects can be rescued by inhibition or knockdown of ROCK. Together, these data suggest that loss of the CCM proteins causes a cytoskeletal defect resulting in a loss of vascular integrity in endothelial cells. However, further definition of the pathophysiology of CCM in patients is limited by the lack of a robust human endothelial cell model that could be used to study CCM biology in the context of actual human ccm mutation. Endothelial progenitor-derived endothelial cells (EP-ECs) allow, for the first time, the ability to study endothelial cells from CCM patients. EP-ECs are easily isolated from whole blood, proliferate to 10^19 cells from a single donor, and produce reliable phenotypes in endothelial functional assays. The aims of this proposal will characterize the critical CCM signaling axis involving RhoA, ROCK, LIM Kinase (LIMK), and Myosin Light Chain 2 (MLC2) and the functional effects of ccm mutation in EP-ECs from familial CCM patients. Aim 1 will define the functional effects of CCM mutation in a heterozygous background with one mutant and one wild type allele, as is found in familial patients. RNAi-mediated loss of heterozygosity by knockdown of the wild-type allele will then be used to define the pathophysiology of bi-allelic loss of protein, effectively mimicking the CCM genotype. Aim 2 will define the RhoA-ROCK-LIMK and -MLC2 signaling axis in EP-ECs from patients with ccm1, -2, or -3 mutation before and after knockdown of the wild type allele. The aims will define regulation of the actin cytoskeleton by the CCM proteins and the ability of therapeutic targeting of the ROCK signaling axis to rescue CCM-related pathophysiology in EP-ECs derived from human patients.